maintenance/convertLinks.php

   1 <?php
   2 /**
   3  * Convert from the old links schema (string->ID) to the new schema (ID->ID)
   4  * The wiki should be put into read-only mode while this script executes
   5  *
   6  * This program is free software; you can redistribute it and/or modify
   7  * it under the terms of the GNU General Public License as published by
   8  * the Free Software Foundation; either version 2 of the License, or
   9  * (at your option) any later version.
  10  *
  11  * This program is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License along
  17  * with this program; if not, write to the Free Software Foundation, Inc.,
  18  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  19  * http://www.gnu.org/copyleft/gpl.html
  20  *
  21  * @file
  22  * @ingroup Maintenance
  23  */
  24
  25 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
  26
  27 class ConvertLinks extends Maintenance {
  28         private $logPerformance;
  29
  30         public function __construct() {
  31                 parent::__construct();
  32                 $this->mDescription = "Convert from the old links schema (string->ID) to the new schema (ID->ID)
  33 The wiki should be put into read-only mode while this script executes";
  34
  35                 $this->addArg( 'logperformance', "Log performance to perfLogFilename.", false );
  36                 $this->addArg( 'perfLogFilename', "Filename where performance is logged if --logperformance was set (defaults to 'convLinksPerf.txt').", false );
  37                 $this->addArg( 'keep-links-table', "Don't overwrite the old links table with the new one, leave the new table at links_temp.", false );
  38                 $this->addArg( 'nokeys', "Don't create keys, and so allow duplicates in the new links table.\n
  39 This gives a huge speed improvement for very large links tables which are MyISAM." /* (What about InnoDB?) */, false );
  40         }
  41
  42         public function getDbType() {
  43                 return Maintenance::DB_ADMIN;
  44         }
  45
  46         public function execute() {
  47                 $dbw = wfGetDB( DB_MASTER );
  48
  49                 $type = $dbw->getType();
  50                 if ( $type != 'mysql' ) {
  51                         $this->output( "Link table conversion not necessary for $type\n" );
  52                         return;
  53                 }
  54
  55                 global $wgContLang;
  56
  57                 $numBadLinks = $curRowsRead = 0; # counters etc
  58                 $totalTuplesInserted = 0; # total tuples INSERTed into links_temp
  59
  60                 $reportCurReadProgress = true; # whether or not to give progress reports while reading IDs from cur table
  61                 $curReadReportInterval = 1000; # number of rows between progress reports
  62
  63                 $reportLinksConvProgress = true; # whether or not to give progress reports during conversion
  64                 $linksConvInsertInterval = 1000; # number of rows per INSERT
  65
  66                 $initialRowOffset = 0;
  67                 # $finalRowOffset = 0; # not used yet; highest row number from links table to process
  68
  69                 $overwriteLinksTable = !$this->hasOption( 'keep-links-table' );
  70                 $noKeys = $this->hasOption( 'noKeys' );
  71                 $this->logPerformance = $this->hasOption( 'logperformance' );
  72                 $perfLogFilename = $this->getArg( 'perfLogFilename', "convLinksPerf.txt" );
  73
  74                 # --------------------------------------------------------------------
  75
  76                 list ( $cur, $links, $links_temp, $links_backup ) = $dbw->tableNamesN( 'cur', 'links', 'links_temp', 'links_backup' );
  77
  78                 if( $dbw->tableExists( 'pagelinks' ) ) {
  79                         $this->output( "...have pagelinks; skipping old links table updates\n" );
  80                         return;
  81                 }
  82
  83                 $res = $dbw->query( "SELECT l_from FROM $links LIMIT 1" );
  84                 if ( $dbw->fieldType( $res, 0 ) == "int" ) {
  85                         $this->output( "Schema already converted\n" );
  86                         return;
  87                 }
  88
  89                 $res = $dbw->query( "SELECT COUNT(*) AS count FROM $links" );
  90                 $row = $dbw->fetchObject( $res );
  91                 $numRows = $row->count;
  92                 $dbw->freeResult( $res );
  93
  94                 if ( $numRows == 0 ) {
  95                         $this->output( "Updating schema (no rows to convert)...\n" );
  96                         $this->createTempTable();
  97                 } else {
  98                         $fh = false;
  99                         if ( $this->logPerformance ) {
 100                                 $fh = fopen ( $perfLogFilename, "w" );
 101                                 if ( !$fh ) {
 102                                         $this->error( "Couldn't open $perfLogFilename" );
 103                                         $this->logPerformance = false;
 104                                 }
 105                         }
 106                         $baseTime = $startTime = $this->getMicroTime();
 107                         # Create a title -> cur_id map
 108                         $this->output( "Loading IDs from $cur table...\n" );
 109                         $this->performanceLog ( $fh, "Reading $numRows rows from cur table...\n" );
 110                         $this->performanceLog ( $fh, "rows read vs seconds elapsed:\n" );
 111
 112                         $dbw->bufferResults( false );
 113                         $res = $dbw->query( "SELECT cur_namespace,cur_title,cur_id FROM $cur" );
 114                         $ids = array();
 115
 116                         foreach ( $res as $row ) {
 117                                 $title = $row->cur_title;
 118                                 if ( $row->cur_namespace ) {
 119                                         $title = $wgContLang->getNsText( $row->cur_namespace ) . ":$title";
 120                                 }
 121                                 $ids[$title] = $row->cur_id;
 122                                 $curRowsRead++;
 123                                 if ( $reportCurReadProgress ) {
 124                                         if ( ( $curRowsRead % $curReadReportInterval ) == 0 ) {
 125                                                 $this->performanceLog( $fh, $curRowsRead . " " . ( $this->getMicroTime() - $baseTime ) . "\n" );
 126                                                 $this->output( "\t$curRowsRead rows of $cur table read.\n" );
 127                                         }
 128                                 }
 129                         }
 130                         $dbw->freeResult( $res );
 131                         $dbw->bufferResults( true );
 132                         $this->output( "Finished loading IDs.\n\n" );
 133                         $this->performanceLog( $fh, "Took " . ( $this->getMicroTime() - $baseTime ) . " seconds to load IDs.\n\n" );
 134
 135                         # --------------------------------------------------------------------
 136
 137                         # Now, step through the links table (in chunks of $linksConvInsertInterval rows),
 138                         # convert, and write to the new table.
 139                         $this->createTempTable();
 140                         $this->performanceLog( $fh, "Resetting timer.\n\n" );
 141                         $baseTime = $this->getMicroTime();
 142                         $this->output( "Processing $numRows rows from $links table...\n" );
 143                         $this->performanceLog( $fh, "Processing $numRows rows from $links table...\n" );
 144                         $this->performanceLog( $fh, "rows inserted vs seconds elapsed:\n" );
 145
 146                         for ( $rowOffset = $initialRowOffset; $rowOffset < $numRows; $rowOffset += $linksConvInsertInterval ) {
 147                                 $sqlRead = "SELECT * FROM $links ";
 148                                 $sqlRead = $dbw->limitResult( $sqlRead, $linksConvInsertInterval, $rowOffset );
 149                                 $res = $dbw->query( $sqlRead );
 150                                 if ( $noKeys ) {
 151                                         $sqlWrite = array( "INSERT INTO $links_temp (l_from,l_to) VALUES " );
 152                                 } else {
 153                                         $sqlWrite = array( "INSERT IGNORE INTO $links_temp (l_from,l_to) VALUES " );
 154                                 }
 155
 156                                 $tuplesAdded = 0; # no tuples added to INSERT yet
 157                                 foreach ( $res as $row ) {
 158                                         $fromTitle = $row->l_from;
 159                                         if ( array_key_exists( $fromTitle, $ids ) ) { # valid title
 160                                                 $from = $ids[$fromTitle];
 161                                                 $to = $row->l_to;
 162                                                 if ( $tuplesAdded != 0 ) {
 163                                                         $sqlWrite[] = ",";
 164                                                 }
 165                                                 $sqlWrite[] = "($from,$to)";
 166                                                 $tuplesAdded++;
 167                                         } else { # invalid title
 168                                                 $numBadLinks++;
 169                                         }
 170                                 }
 171                                 $dbw->freeResult( $res );
 172                                 # $this->output( "rowOffset: $rowOffset\ttuplesAdded: $tuplesAdded\tnumBadLinks: $numBadLinks\n" );
 173                                 if ( $tuplesAdded != 0  ) {
 174                                         if ( $reportLinksConvProgress ) {
 175                                                 $this->output( "Inserting $tuplesAdded tuples into $links_temp..." );
 176                                         }
 177                                         $dbw->query( implode( "", $sqlWrite ) );
 178                                         $totalTuplesInserted += $tuplesAdded;
 179                                         if ( $reportLinksConvProgress )
 180                                                 $this->output( " done. Total $totalTuplesInserted tuples inserted.\n" );
 181                                                 $this->performanceLog( $fh, $totalTuplesInserted . " " . ( $this->getMicroTime() - $baseTime ) . "\n"  );
 182                                 }
 183                         }
 184                         $this->output( "$totalTuplesInserted valid titles and $numBadLinks invalid titles were processed.\n\n" );
 185                         $this->performanceLog( $fh, "$totalTuplesInserted valid titles and $numBadLinks invalid titles were processed.\n" );
 186                         $this->performanceLog( $fh, "Total execution time: " . ( $this->getMicroTime() - $startTime ) . " seconds.\n" );
 187                         if ( $this->logPerformance ) {
 188                                 fclose ( $fh );
 189                         }
 190                 }
 191                 # --------------------------------------------------------------------
 192
 193                 if ( $overwriteLinksTable ) {
 194                         # Check for existing links_backup, and delete it if it exists.
 195                         $this->output( "Dropping backup links table if it exists..." );
 196                         $dbw->query( "DROP TABLE IF EXISTS $links_backup", __METHOD__ );
 197                         $this->output( " done.\n" );
 198
 199                         # Swap in the new table, and move old links table to links_backup
 200                         $this->output( "Swapping tables '$links' to '$links_backup'; '$links_temp' to '$links'..." );
 201                         $dbw->query( "RENAME TABLE links TO $links_backup, $links_temp TO $links", __METHOD__ );
 202                         $this->output( " done.\n\n" );
 203
 204                         $dbw->close();
 205                         $this->output( "Conversion complete. The old table remains at $links_backup;\n" );
 206                         $this->output( "delete at your leisure.\n" );
 207                 } else {
 208                         $this->output( "Conversion complete.  The converted table is at $links_temp;\n" );
 209                         $this->output( "the original links table is unchanged.\n" );
 210                 }
 211         }
 212
 213         private function createTempTable() {
 214                 $dbConn = wfGetDB( DB_MASTER );
 215
 216                 if ( !( $dbConn->isOpen() ) ) {
 217                         $this->output( "Opening connection to database failed.\n" );
 218                         return;
 219                 }
 220                 $links_temp = $dbConn->tableName( 'links_temp' );
 221
 222                 $this->output( "Dropping temporary links table if it exists..." );
 223                 $dbConn->query( "DROP TABLE IF EXISTS $links_temp" );
 224                 $this->output( " done.\n" );
 225
 226                 $this->output( "Creating temporary links table..." );
 227                 if ( $this->hasOption( 'noKeys' ) ) {
 228                         $dbConn->query( "CREATE TABLE $links_temp ( " .
 229                         "l_from int(8) unsigned NOT NULL default '0', " .
 230                         "l_to int(8) unsigned NOT NULL default '0')" );
 231                 } else {
 232                         $dbConn->query( "CREATE TABLE $links_temp ( " .
 233                         "l_from int(8) unsigned NOT NULL default '0', " .
 234                         "l_to int(8) unsigned NOT NULL default '0', " .
 235                         "UNIQUE KEY l_from(l_from,l_to), " .
 236                         "KEY (l_to))" );
 237                 }
 238                 $this->output( " done.\n\n" );
 239         }
 240
 241         private function performanceLog( $fh, $text ) {
 242                 if ( $this->logPerformance ) {
 243                         fwrite( $fh, $text );
 244                 }
 245         }
 246
 247         private function getMicroTime() { # return time in seconds, with microsecond accuracy
 248                 list( $usec, $sec ) = explode( " ", microtime() );
 249                 return ( (float)$usec + (float)$sec );
 250         }
 251 }
 252
 253 $maintClass = "ConvertLinks";
 254 require_once( DO_MAINTENANCE );